Homing flights of pigeons in the Frankfurt region: the effect of distance and local experience

• A meta-analysis of tracks documents the behaviour of homing pigeons.
• Behaviour at release sites reflects how pigeons integrate new sites into the map.
• Pigeons repeatedly homing from the same site continue to choose different routes.
• A comparison with data from the Oxford regions reveals differences in strategy.
• Regional differences in strategy reflect different availability of navigational cues.

In a meta-analysis, we analysed GPS-recorded tracks of homing pigeons, focusing on how initial behaviour and track structure change with increasing distance from the loft and on the effect of local experience. ‘Points of Decision’ divide the tracks into several phases, with the first marking the end of the initial phase, when the pigeon begins to leave the release site. With increasing distance, the number of Points of Decision increases significantly, whether pigeons are unfamiliar or familiar with the release site. In unfamiliar birds, the duration of the initial phase increases with increasing distance from the home loft, while agreement among birds decreases. These correlations with distance, not found in familiar birds, appear to reflect the necessity to interpret the still unfamiliar combination of ‘map’ factors. Pigeons starting a second time from a site take more time to leave and initially fly over a larger area, which we interpret as a procedure to integrate the respective site into their ‘map’. When pigeons are released repeatedly from the same sites, the behaviour at the beginning of the flight does not change consistently, but there are fewer decision points and overall efficiency increases. The flight routes chosen by individuals continue to differ; route stereotypies, as described in the Oxford region, are not observed. This suggests loft-specific difference in the weighting and rating of navigational factors, probably caused by the availability and suitability of these factors in the respective regions. Our findings warn against simply transferring findings from one loft to another.